Parameters estimation of squirrel-cage induction motors using ANN and ANFIS

Parameters estimation; Induction motor; Squirrel-cage; Artificial neural network; Adaptive neuro-fuzzy inference system Abstract In the transient behavior analysis of a squirrel-cage induction motor, the parameters of the single-cage and double-cage models are studied. These parameters are usually hard to obtain. This paper presents two new methods to predict the induction motor parameters in the singlecage and double-cage models based on artificial neural network (ANN) and adaptive neurofuzzy inference system (ANFIS). For this purpose, the experimental data (manufacturer data) of 20 induction motors with the different power are used. The experimental data are including of the starting torque and current, maximum torque, full load sleep, efficiency, rated active power and reactive power. The obtained results from the proposed ANN and ANFIS models are compared with each other and with the experimental data, which show a good agreement between the predicted values and the experimental data. But the proposed ANFIS model is more accurate than the proposed ANN model. 2016 Faculty of Engineering, Alexandria University. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).